362 FERTILISATION OF THE GOODENIACE.^, 



stages, which are deciduous), and rarely ciliate on the lip. The 

 lip of the indusium is shallow, and usually divided into two, or 

 sometimes four, by notches. The indusium and stigma are in 

 most species of a very dark colour. The stigma rarely grows out 

 to project beyond the indusium lips, which is also the case in 

 other plants of the order. The j)lants, with the exception of D, 

 diversifolia, De Vr., are always clothed with silky, cottony, woolly, 

 stellate or branching hairs on the calyx, and the outside of the 

 corolla (excejDt the wings) and sometimes on the stems and leaves. 

 The flowers are almost always blue or purple. In the remarks on 

 the genus in Flora Australiensis [1] the colour is said to be blue, 

 purple, red, white, or rarely 3^ellow. This latter colour I have 

 not seen in any fresh specimens (except in the tube), and in dried 

 plants it is very difficult to make out the colour. But from an 

 analysis of the descriptions in Flora Australiensis, it appears that 

 of the 34 species described, 23 are blue or purple, 1 white, and 1 

 (D. rosmarinifolia, Schl.) is said to be white, blue or red; while 9 

 have no colour mentioned. 



The characteristic features by which they may be recognised 

 are the solitary ovules, connate anthers, the auricles, and the 

 hairless style and indusium. 



From their dense covering of hairs they have the aspect of 

 desert plants, and judging by the number of species collected by 

 the Elder Expedition, they are plentiful in the arid interior of 

 Australia as compared with other genera of the order. And 

 many of the species of other genera occurring there are alsa 

 tomentose. The Census of Australian Plants [2] gives in all 38 

 species of the genus, which are distributed as follows : — West 

 Australia has 29 species, 26 of Avhich are endemic; South Australia 

 5 species, none endemic (this number would probably be much 

 higher were the central parts of the colony collected over); 

 Victoria has 5 species, none endemic; Tasmania has 1 species 

 found elsewhere also; New South Wales has 7 species, 1 being 

 confined to the colon}^; Queensland has 6 species, 2 endemic; 

 and North Australia 1 species, o ?curring elsewhere also. From 

 this it will be seen that the head-quarters of the genus is in West 



